A constructive no-confidence vote is per implication secret
- Loammi Wolf
- 15 May 2017 12:28 (South Africa)
A report in Business Day noted that it is not likely that the Constitutional Court would “force a secret ballot on Parliament” over the no-confidence motion in Zuma. The views expressed in the article might actually resonate more with the practice under the previous Westminster constitutions and do not take the rules pertaining to a constructive vote of no confidence into account.
Of course, nobody can predict how the Constitutional Court will rule, but there are good reasons to overturn the precedent in the Western Cape High Court’s ruling in the Tlouamma matter.
The gist of the argument that was also presented in the Tlouamma case is that the separation of powers demands that Parliament has the sole prerogative to determine its own procedures and exercise powers designated to it by the Constitution.
In the current matter before the Constitutional Court there is no dispute that the dismissal of the Cabinet including the president with a no-confidence vote falls in the exclusive domain of powers of the National Assembly. It is a matter of constitutional interpretation, however, as to whether the Constitution requires that the vote should be by secret ballot.
The arguments favouring an open ballot to guarantee transparency of how Parliament exercises its powers do not offer any explanation as to why the Constitution then explicitly prescribes a secret ballot for democratic legitimation procedures.
It is common cause that the Constitution prescribes a secret ballot for the democratic legitimation of specific constitutional office-bearers in Schedule 3A. Their number includes the president, the speaker and the deputy speaker of the National Assembly, the chairperson and deputy chairperson of the Council of Provinces, and the provincial premiers. The procedure requires that nominated candidates must sign the nomination form and that the election takes place without any debate in Parliament. The winning candidate is appointed by majority vote in a secret ballot.
The National Assembly exercises the constituent power to legitimise the head of the executive to form a government. It can also withdraw the confidence placed in the president with a constructive vote of no confidence.
This legitimation procedure is fundamentally different from that in Westminster systems where the head of state, be that a constitutional monarch like in the United Kingdom or a state president as under the 1910 and 1961 Constitutions of South Africa, tasks the leader of the strongest party in Parliament to form a government. In the United Kingdom, one also calls the cabinet “Her Majesty’s government”. In South Africa it was common to say that the Cabinet “serves at the pleasure” of the state president. It is important to note that even in those days the Cabinet did not serve “at the pleasure of the prime minister” as head of government.
The 1983 Constitution merged the positions of head of state and head of government with the effect that it became really absurd to maintain that that the head of state should legitimise himself in the dual capacity as the head of government to form a government. It is equally ridiculous to claim that he and his Sabinet served “at the pleasure” of the head of state, because it is one and the same person.
The 1993 Interim Constitution and the 1996 Constitution kept the merged office of head of state and head of government, but changed the way of legitimising the executive. A new system was established in terms of which the executive is legitimised by the National Assembly and not the head of state. This shift is embedded in the notion of popular sovereignty where all state power emanates from the people (section 42(3) of the Constitution). Thus, the elected representatives mandate a president to form a government.
Section 86(1) of the 1996 Constitution determines that the members of the Assembly elect “a woman or a man from among its members to be the President”. The rationale behind this provision is that the government derives its legitimacy solely from the Members of Parliament that elect the head of government on the basis of the dual principles of democratic representation and collegiality.
Section 86(2) of the Constitution further specifies that Parliament elects the president by majority vote in a secret ballot if more than one candidate is nominated. This provision is similar to article 67 of the German Constitution: the federal parliament also elects the chancellor by majority vote in a secret ballot.
The South African president, like his German counterpart, is elected by all Members of Parliament. He or she does not only represent the strongest party in Parliament as leader of the government but the Assembly as elected representatives of people.
The secret ballot requirement has good reasons and serves a twofold purpose. First, it guarantees that every member has a free choice to cast a vote without intimidation or undue influence, and second, it safeguards collegiality insofar as it avoids a situation where candidates could harbour hard feelings when certain members did not vote for them. That would hamper the work of the legislature and the executive during the legislative term.
A second difference compared to the Westminster system is how a no-confidence vote is regulated and what effect it has. Most Westminster systems are characterised by a two-party system. Coalition governments are extremely rare compared to countries with proportional electoral systems. A majority party will therefore seldom lose power with a vote of no confidence because the members of the majority party will usually rally behind its government and defeat a vote of no confidence. In the United Kingdom, there have been a total of 11 prime ministers defeated through a no-confidence motion, and only one (James Callaghan) since 1925.
In countries with a Westminster constitutional system, a vote of no confidence is traditionally directed against the entire cabinet. When the motion is passed by majority vote, the government comes to a fall and elections must be called to get a new mandate from the voters. In the United Kingdom, both a successful no-confidence vote and a lost confidence vote trigger general elections. The difference between the two kinds of motions is that the no-confidence vote is a weapon of the parliament tabled by the opposition, and the confidence vote is a pressure mechanism of the head of government to get parliament to endorse specific policy measures.
The South African and German constitutions are different in this respect too.
Section 102 of the South African Constitution provides two options for a motion of no confidence: first, the motion could only be in the Cabinet excluding the president, and second, in the president. In the first instance, the president must reconstitute the Cabinet and can continue to govern for the residual part of the legislative term. In the second case, the president and the entire Cabinet must resign.
A successful no-confidence vote in either of these two constellations does not automatically bring the government to a fall with the effect that a general election must be called like in Westminster systems.
The effect of a successful no-confidence vote in the president is that it creates a vacancy in the office of the president, which should be filled according to the procedures prescribed by section 86(3) of the Constitution. When the president and the Cabinet are forced to resign, the Assembly must by secret ballot elect a new president within 30 days. The candidate with an absolute majority then becomes president and can form a new government for the rest of the legislative term. If no candidate gets an absolute majority, the acting president must dissolve the Assembly and call general elections (section 55(2) of the Constitution).
This type of no confidence vote, which allows a reconstitution of the executive and does not automatically trigger a general election, is referred to as a “constructive vote of no confidence”.
Article 67 of the German Constitution also provides for a constructive vote of no confidence, but is different from the South African variation in two respects: First, it does not cater for an option where Parliament can pass a motion of no-confidence only in the cabinet excluding the chancellor, and second, the replacement of the challenged chancellor must be elected on the same ballot as the no-confidence vote. The members of parliament (usually the opposition) must, on the same ballot that proposes a no-confidence motion in the chancellor, propose another candidate whom they want to be appointed as successor of the chancellor. A motion of no confidence must therefore take two hurdles to be successful: first, a majority of the members of parliament must vote in favour of the motion, and second, the alternative candidate must be endorsed by a majority.
This only happened once since the 1949 Constitution was adopted. In October 1982, the German parliament tabled a motion of no confidence in Chancellor Helmut Schmidt when the coalition government of the SPD (Social Democrats) and the FDP (Free Democrats) broke up due to protests about the Nato Double-Track Decision on the stationing of medium-range ballistic missiles in Western Europe.The FDP then sided with the CDU (Christian Democrats) and put up Helmut Kohl as their candidate. The no confidence vote in Schmidt was carried, Kohl got an absolute majority and formed a new government.
Thus, in both countries the constructive vote of no confidence requires that the vacancy in the office of the head of government after a successful no-confidence vote should be filled for the rest of the legislative period. The two phases of the procedure to achieve this are intrinsically linked. The only difference is that in Germany the replacement must be elected on the same ballot and in South Africa within 30 days after the successful vote to dismiss the head of government from office.
The rationale for this kind of construct is to avoid many short-lived governments in systems where coalition governments are the norm and not the exception. It creates more political stability.
Section 102 of the South African Constitution and article 67 of the German Constitution both require that a no-confidence vote should be “by majority vote”, but does not explicitly state that the vote must be by secret ballot.
In Germany, the decision is nevertheless taken by secret ballot because the type of voting which is required is a ballot to replace the head of government. The members of parliament are not allowed to vote by raising their hands or by electronic voting like in ordinary parliamentary procedures. Ordinary voting procedures are never secret because the constitution requires transparency, but legitimation procedures are an exception to the rule. The rationale behind the secret ballot in a no confidence vote is the same as the rationale for a secret ballot in the election of the head of government. The withdrawing of confidence is the mirror-image procedure of the confidence vested in the head of government upon election of the head of government.
In the case of the no-confidence vote in Schmidt, not only the FDP and CDU members voted against him but apparently also members of his own party. For many years he wondered who they could have been. He felt hurt, but could not hold it against anybody, which certainly was a good thing to avoid bad blood.
The difference between the secret and open ballots is that the caucus of a party can expect its members to vote according to the party line in order to put the ruling party’s or coalition government’s policy into practice.
One must therefore distinguish the democratic legitimation procedures of constitutional office-bearers from the ordinary business of Parliament. Thus, section 53 of the South African Constitution determines that “except where the Constitution provides otherwise”, decisions in Parliament are taken by an ordinary majority of the votes cast. It prescribes a quorum of a majority of the members of the Assembly (ie 50%) to vote on bills, and at least a third of the members on any other question before the Assembly. The legitimation procedures and no confidence motions where the confidence in the government is withdrawn do not fall in this category: here a majority of all the members is required (ie 50% plus one).
Section 57(1) of the Constitution refers to the making of rules by the Assembly to “determine and control its internal arrangements, proceedings and procedures”. Such rules and orders should have due regard to accountability and transparency. In this spirit section 59(1)(b) of the Constitution also requires that the Assembly must “conduct its business in an open manner” and hold its sittings in public insofar as it is reasonably possible.
The question is what exactly such “internal proceedings” would encompass. It would obviously include the drafting of bills and legislative deliberations or debates on matters of public concern. The difficulty is that legitimation procedures also fall in the exclusive domain of powers of the Assembly: the election of the speaker and deputy speaker would obviously be an “internal” matter; yet, the election of the president can hardly be regarded as an “external” matter.
The “internal” requirement as such is therefore not really helpful to arrive at a conclusion on when open ballots would be appropriate or not. The key is to be found in section 53 that states “except where the Constitution provides otherwise”. In legitimation procedures other constitutional principles take precedence. Here transparency is not the paramount consideration: a free democratic choice without intimidation and undue influence and the principle of collegiality overrides the general principle of an open ballot to secure transparency. The act of legitimising the government through the election of the president to form a government cannot be disconnected from acts that de-legitimise the government by withdrawing the confidence put in the government. DM
Loammi Wolf is a constitutional law specialist who runs the initiative Democracy for Peace.